This dissertation explores co-construction of a situated activity in a current unit by mainly focusing on hearers' actions. The "co-construction" involves a process of interaction in which speakers and hearers jointly construct the form and the meaning of ongoing utterances even when they disagree with each other (Jacoby & Ochs, 1995). In particular, this study addresses (1) interactional resources that Korean speakers employ to project their talk and action during the progress of talk, (2) some ways in which hearers display their alignment to a situated activity emerging from the current interaction, and (3) possible consequences of recipients' action to the syntactic structure of ongoing talk. The issue of this dissertation touches on controversies among previous studies, with respect to the relationship between interaction and linguistic typology. The majority of studies claim that the projection of talk and action in SOV languages is "delayed" due to the predicate-final structures and agglutination in contrast to SOV languages (Fox et al., 1996, Tanaka, 1999, 2000; Hayashi, 2004 inter alia). That is, recipients in Korean and other SOV languages must "wait and see" to recognize the trajectory of speakers' talk. However, recent studies suggest that recipients can collaboratively construct a speaker's emerging talk at an early stage, even in such SOV languages as Japanese (Hayashi, 2003; S. Iwasaki, 2008, 2009). This study builds on previous literature by examining how Korean speakers and recipients closely monitor each other's talk and action to construct a current unit of talk collaboratively. While acknowledging the structural implication of delayed projectability, this study pays central attention to ways in which co-construction is afforded as early as possible in the process of emerging talk. In particular, this study examines the use of participants' nonverbal actions including gestures as interactional resources to achieve co-construction. Moreover, this study shows how participants use grammar as resources during ongoing interaction to construct talk and action. The results of this study suggest that co-construction of a unit and a situated activity is accomplished through early projection of talk (and action), which is made possible through multimodal resources including gestures and structural adaptation in Korean interaction. By demonstrating how Korean participants take into account multiple kinds of resources to make sense of each other's talk and action even prior to the occurrences of a particle or a predicate, this study sheds light on the interface between language and interaction. In particular, the results of this study demonstrate that in even SOV languages, early projection of units may be possible when we consider multimodal resources (see S. Iwasaki, 2008, 2009). [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]